Abstract
The ability to actively maintain recently viewed visual information is essential to many aspects of daily life, and working memory for faces is of particular importance for human social interactions. Recent studies have found that working memory performance is better for upright faces than inverted faces, but it remains unclear whether this advantage for upright faces is attributable to superior memory capacity for these highly familiar stimuli (Curby & Gauthier, 2007) or to the superior precision with which these items can be retained (Scolari, Vogel, & Awh, 2008). Here, we investigated whether upright faces are retained in working memory with superior precision or capacity by assessing participants’ reported memories of such stimuli, using a continuously varying set of computer-generated faces. Observers viewed 1-5 sequentially presented faces for 1.5s each, then after a 3s retention interval, a spatial cue indicated which face to report from memory via method of adjustment. Faces were presented upside-down on alternating blocks of trials. We analyzed the distribution of errors using a mixed model approach to separately estimate the precision of memory for successfully retained faces and the frequency of random guess responses (Zhang and Luck, 2008). Pooled results from 16 participants indicated that specialized processing of upright faces conferred an advantage to both working memory capacity and resolution. Estimated capacities for upright faces were greater than for inverted faces (~3.5 vs. 2.5, respectively), and the precision of working memory also proved superior in the upright condition. Taken together, our results suggest that upright faces can be represented more efficiently in working memory, allowing for greater capacity, while at the same time these upright faces can be maintained with greater visual precision.
Meeting abstract presented at VSS 2012